基于逆向云模型的雷达目标识别  

Radar target recognition based on the backward cloud model

作　　者：吴杰[1] 周建江[1] 朱劼昊[1] 

机构地区：[1]南京航空航天大学信息科学与技术学院,江苏南京210016

出　　处：《光电子．激光》2010年第10期1556-1559,共4页Journal of Optoelectronics·Laser

基　　金：装备预研重点基金资助项目(N0601-041);中电集团第14研究所院士基金资助项目(2008041001)

摘　　要：提出了一种基于逆向云模型的雷达目标识别方法。首先,基于模-1距离准则建立了云滴确定度,对高分辨距离像(HRRP)距离单元进行正态逆向云模型建模;然后,定义逆云隶属度表征待测样本属于某个训练类别的概率。对目标模板库中5种飞机高分辨距离像数据的仿真结果表明,该方法具有识别率高、对训练样本量的要求较为宽松和方位角划分不敏感等优点,是一种有效的雷达目标识别方法。A radar target recognition method based on backward cloud model is proposed in this paper.Firstly the certainty degree of cloud drop is established based on the mode-1 criteria,and then the normal backward cloud model is established for the high-resolution range profile（HRRP） unit.Meanwhile,the backward cloud membership is defined to characterize the probability of the test sample belonging to a training class.Simulation results based on the HRRP dataset of five aircraft models demonstrate that the proposed approach is effective for radar target identification,and it has the advantages of high recognition rate,more relaxed requirements for the training sample and insensitivity to azimuth division.

关 键 词：雷达自动目标识别 高分辨距离像(HRRP) 逆云模型 逆云隶属度 

分 类 号：TN911.7[电子电信—通信与信息系统]